First Draft Risk Profile For

1 2 Stockholm Convention on Persistent Organic Pollutants

3 POPs Review Committee (POPRC)

7 ENDOSULFAN

9 THIRD DRAFT RISK PROFILE

14 Draft prepared by the ad hoc working group on Endosulfan 15 under the POPs Review Committee 16 of the Stockholm Convention 17 18 19 20 21 22

23 June 2009

24 25 26

2 1Endosulfan third draft risk profile, June 2009

2 1 Table of Contents 2 3 Executive summary...... 5 4 1. Introduction...... 7 5 1.1 Chemical identity...... 7 6 1.2 Conclusion of the Review Committee regarding Annex D information...... 9 7 1.3 Data sources...... 9 8 1.4 Status of the chemical under international conventions...... 10 9 2. Summary information relevant to the risk profile...... 11 10 2.1 Sources...... 11 11 2.1.1 Production, trade, stockpiles...... 11 12 2.1.2 Uses...... 11 13 2.1.3 Releases to the environment...... 12 14 2.2 Environmental fate...... 12 15 2.2.1 Persistence...... 12 16 2.2.2 Bioaccumulation...... 15 17 2.2.3 Potential for long-range environmental transport...... 18 18 2.3 Exposure...... 20 19 2.3.1 Environmental monitoring data...... 20 20 2.4 Hazard assessment for endpoints of concern...... 30 21 3. Synthesis of information...... 34 22 4. Concluding statement...... 39 23 5 References...... 40

32 1 Endosulfan third draft risk profile, June 2009

1 Executive summary 2Endosulfan is a synthetic organochlorine compound consisting to two isomers (α and β). It is commonly used 3as an agricultural insecticide. It has been sold from the mid 1950s but it is still contained in pesticide products 4in some countries worldwide. Technical endosulfan is a 2:1 to 7:3 mixture of the α- and β-isomers.

5Endosulfan is now banned in at least 60 countries with former uses replaced and its production is decreasing. 6However, endosulfan is still used in different regions of the world.

7Endosulfan aerobic transformation occurs via biologically mediated oxidation. The main metabolite formed is 8endosulfan sulfate. This compound is slowly degraded to the more polar metabolites endosulfan diol, 9endosulfan lactone, endosulfan ether. The combined median dissipation half-life DT50 measured in laboratory 10studies for α and β endosulfan and endosulfan sulfate, was selected as a relevant parameter for quantifying the 11persistence, it ranges typically between 28 and 391 days. In the aquatic compartment, endosulfan is stable to 12photolysis; a rapid hydrolysis is only observed at high pH values, and it is non-readily biodegradable. In 13water/sediment systems, DT50 > 120 d was demonstrated. There is a uncertainty on the degradation rate of 14endosulfan in the atmosphere, however it is expected that the half life exceeds the 2 days threshold.

15The bioconcentration potential of endosulfan in aquatic organisms is confirmed by experimental data. The 16validated bioconcentration factor (BCF) values range between 1000 and 3000 for fish, from 12 to 600 for 17aquatic invertebrates; and up to 3278 in algae. Thus, reported BCFs are below the criterion of 5,000; and the 18log Kow is measured at 4.7, which is below the criterion of 5. However, modelling suggests that endosulfan 19has an inherent high biomagnification potential, associated with its high K oa, in air-breathing organisms of 20terrestrial, marine mammalian, and human food chains. Additionally, endosulfan was detected in adipose 21tissue and blood of animals in the Arctic and the Antarctic. Endosulfan has also been detected in the blubber 22of minke whales and in the liver of northern fulmars. Therefore, there is sufficient evidence that endosulfan 23enters the food chain and that it bioaccumulates and biomagnifies in the food webs.

24The potential of endosulfan for long range transport (LRT) has been confirmed from three main information 25sources: the analysis of the endosulfan properties, the application of LRT models, and the review of existing 26monitoring data in remote areas.

27LRT has been confirmed by the presence of endosulfan in air and biota from remote areas. Most studies 28measure α- and β-endosulfan, and in some cases, endosulfan sulfate. Other endosulfan metabolites are only 29rarely quantified. The presence of endosulfan in remote areas, far away from intensive use areas, in particular, 30the Arctic and Antarctica has been confirmed. The potential for LRT, seems to be mostly related to 31atmospheric transfer; deposition at high altitude mountain areas has been also observed..

32The toxicity and ecotoxicity of endosulfan is well documented. Endosulfan is highly toxic for humans and for 33most animal taxa, showing both acute and chronic effects at relatively low exposure levels. Acute lethal 34poisoning in humans and clear environmental effects on aquatic and terrestrial communities has been 35observed under standard use conditions when the risk mitigation measures have not been followed. Several 36countries have found that endosulfan poses unacceptable risks, or has caused unacceptable harm, to human 37health and the environment, and have banned or severely restricted it.

39Finally, the role of endosulfan metabolites other than endosulfan sulfate has received limited attention. 40Endosulfan lactone has the same chronic NOEC value as the parent endosulfan isomers. If the toxicity of each 41metabolite is integrated into the degradation/metabolism process, the result is a biphasic curve, the initial 42degradation step, to endosulfan sulfate, increases the bioaccumulation potential and maintains or slightly 43reduces the toxicity; the further degradation steps reduce the toxicity and bioaccumulation potential, but then 44further steps, with the formation of the lactone, increase again the toxicity and the bioaccumulation potential.

3 3 1Endosulfan third draft risk profile, June 2009

2 1Based on the inherent properties, and given the widespread occurrence of endosulfan in environmental 2compartments and biota in remote areas which compared to toxicological values suggest a significant risk for 3adverse effects, together with the uncertainty associated with the insufficiently understood role of the 4metabolites which maintain the endosulfan chemical structure, it is concluded that endosulfan is likely, as a 5result of its long-range environmental transport, to lead to significant adverse human health and 6environmental effects, such that global action is warranted.

71. Introduction

8Endosulfan is a synthetic organochlorine compound. It is commonly used as an agricultural insecticide. It has 9been sold from the mid 1950s but it is still contained in pesticide products in some countries worldwide. 10Technical information about (eco)toxicity, environmental fate, residues in food and feedstuff, environmental 11concentrations, etc. of endosulfan is widely available from different sources around the world. Various 12reviews have been published during the last decade regarding every aspect related to our environment.

13 1.1 Chemical identity

14Names and registry numbers

Common name endosulfan IUPAC Chem. 6,7,8,9,10,10-hexachloro-1,5,5a,6,9,9a-hexahydro-6,9-methano-2,4,3- Abstracts benzodioxathiepin-3-oxide 6,9-methano-2,4,3-benzodioxathiepin-6,7,8,9,10,10-hexachloro-1,5,5a,6,9,9- hexahydro-3-oxide CAS registry numbers alpha (α) endosulfan 959-98-8 beta (β) endosulfan 33213-65-9 technical endosulfan * 115-29-7 Endosulfan sulfate: * stereochemically 1031-07-8 unspecified Trade name Thiodan® , Thionex, Endosan, Farmoz, Endosulfan, Callisulfan 15* Technical endosulfan is a 2:1 to 7:3 mixture of the α- and the β-isomer.

16Technical grade endosulfan is a diastereomeric mixture of two biologically active isomers (α- and β-) in 17approximately 2:1 to 7:3 ratio, along with impurities and degradation products. The technical product must 18contain at least 94% endosulfan in accord with specifications of the Food and Agricultural Organization of the 19United Nations (FAO Specification 89/TC/S) with content of the α-isomer in the range of 64-67% and the β- 20isomer of 29-32%. The α-isomer is asymmetric and exists in two twist chair forms while the β-form is 21symmetric. The β-isomer is easily converted to α-endosulfan, but not vice versa (INIA, 1999).

22Structures

Molecular formula C9H6Cl6O3S C9H6Cl6O4S Molecular mass 406.96 g·mol-1 422.96 g·mol-1 Structural formulas of the isomers and the main transformation product

α-endosulfan β-endosulfan endosulfan sulphate

23Physical and chemical properties of endosulfan isomers and of endosulfan sulfate

2 α isomer β isomer Technical sulfate mixed isomers Melting point, ºC 109.2 213.3 70-124 181 - 201 Solubility in water 0.33 0.32 0.05-0.99 0.22 pH 5, at 25ºC, mg/L Recommended value: 0.5 Vapour Pressure, Pa, at 1.05 E-03 1.38 E-04 2.27E-5 – 1.3E-3 2.3 E-05 25ºC Recommended value: 1.3E-3 Henry’s Law Constant 1.1 0.2 1.09-13.2, recommended Pa m3/mol, at 20ºC value: 1.06 logarithm octanol-water 4.7 4.7 3.6 3.77 partition coefficient (Log Kow) at pH 5.1 Dissociation constant n.a. (no acidic n.a. (no acidic n.a. (no acidic protons) n.a. (no acidic protons) protons) protons)

1 1.2 Conclusion of the Review Committee regarding Annex D information

2The Committee evaluated Annex D information at its fourth meeting held in Geneva, from October 13th to 17th 32008, and decided that “it is satisfied that the screening criteria have been fulfilled for endosulfan” and 4concluded that “endosulfan met the screening criteria specified in Annex D”.

5 1.3 Data sources

6The primary source of information for the preparation of this risk profile was the proposal submitted by the 7European Community and its member States that are Parties to the Convention, contained in document 8UNEP/POPS/POPRC.4/14, and additional information submitted for Annex D evaluation. In particular:

9 . INIA 1999-2004. Monograph prepared in the context of the inclusion of the following active 10 substance in Annex I of the Council Directive 91/414/EEC. Instituto Nacional de Investigación y 11 Tecnología Agraria y Alimentaria (I.N.I.A.) including addenda.

12In addition the following parties and observers have answered the request for information specified in Annex 13E of the Convention: Albania, Australia, Bahrain, Bulgaria, Canada, China, Congo (RDC), Costa Rica, 14Croatia, Czech Republic, Ecuador, Egypt, Ghana, Honduras, Japan, Lithuania, Mali, Mauritius, Mexico, New 15Zealand, Nigeria, Norway, Romania, Slovakia, Switzerland, Togo, United States of America, Makteshim- 16Agan Industries (MAI), CropLife, Indian Chemical Council (ICC), Pesticide Action Network (PAN) 17International and the International POPs Elimination Network (IPEN). A more elaborated summary of the 18submissions is provided as separate informal document. Summary of data submitted by Parties and observers 19for information specified in Annex E of the Convention.

20 1.4 Status of the chemical under international conventions

21Endosulfan is subject to a number of regulations and action plans:

22 . In March 2007 the Chemical Review Committee of Rotterdam Convention on the Prior Informed 23 Consent Procedure (PIC) agreed to forward to the conference of the parties of the Convention the 24 recommendation for inclusion of endosulfan in Annex III. Annex III is the list of chemicals that have 25 been banned or severely restricted for health or environmental reasons by parties.

26 . Endosulfan is recognized as one of the twenty-one high-priority compounds identified by PNUMA- 27 GEF (United Nations Environment Programme – Global Environment Facility) during the Regional 28 Evaluation of Persistent Toxic Substances (STP), 2002. These reports have taken into account the 29 magnitude of usage, environmental levels and effects for human beings and for the environment of 30 this compound.

2 1 . The Sahelian Pesticides Committee (CSP) has banned all formulations containing endosulfan. The 2 CSP is the structure for the approval of pesticides for CILSS member States (Burkina Faso, Cap 3 Verde, Chad, Gambia, Guinea Bissau, Mali, Mauritania, Niger and Senegal). The deadline set up for 4 termination of the use of existing stocks of endosulfan was 31/12/2008.

5 . The UN-ECE (United Nations Economic Commission for Europe) has included endosulfan in Annex 6 II of the Draft Protocol on Pollutant Release and Transfer Registers to the AARHUS Convention on 7 access to Information, Public Participation in Decision-making and Access to Justice in 8 Environmental Matters.

9 . The OSPAR has included endosulfan in the List of Chemicals for Priority Action (update 2002)

10 . In the Third North Sea Conference (Annex 1A to the Hague Declaration), endosulfan was agreed on 11 the list of priority substances.

122. Summary information relevant to the risk profile

13 2.1 Sources

142.1.1 Production, trade, stockpiles

15Endosulfan is synthesized via the following steps: Diels-Alder addition of hexachloro-cyclopentadiene and 16cis-butene-1,4-diol in xylene. Reaction of this cis-diol with thionyl chloride forms the final product.

17 Endosulfan was developed in the early 1950s. Global production of endosulfan was estimated to be 10,000 18tonnes annually in 1984. Current production is judged to be significantly higher than in 1984. India is 19regarded as being the world’s largest producer (9900 tonnes per year (Government of India 2001-2007)) and 20exporter (4104 tonnes in 2007-08 to 31 countries (Government of India)); followed by Germany 21(approximately 4000 tonnes per year); China (2400 tonnes), Israel and South Korea.

222.1.2 Uses

23Endosulfan is an insecticide used to control chewing, sucking and boring insects, including aphids, thrips, 24beetles, foliar feeding caterpillars, mites, borers, cutworms, bollworms, bugs, white fliers, leafhoppers, snails 25in rice paddies, earthworms in turf, and tsetse flies.

26Endosulfan is used on a very wide range of crops. Major crops to which it is applied include soy, cotton, rice, 27and tea. Other crops include vegetables, fruit, nuts, berries, grapes, cereals, pulses, corn, oilseeds, potatoes, 28coffee, mushrooms, olives, hops, sorghum, tobacco, and cacao. It is used on ornamentals and forest trees, and 29has been used in the past as an industrial and domestic wood preservative.

30The use of Endosulfan is now banned in at least 60 countries 1 with former uses replaced by less hazardous 31products and methods. More detailed information on current uses as informed by countries is provided as 32separate informal document. Summary of data submitted by Parties and observers for information specified in 33Annex E of the Convention.

31 Austria, Bahrain, Belgium, Belize, Benin, Bulgaria, Burkina Faso, Cambodia, Cape Verde,Chad, Colombia, Cote d’Ivoire, 4Croatia, Cyprus, Czech Republic, Denmark, Egypt, Estonia, Finland, France, Gambia, Germany, Greece, Guinea Bissau, 5Hungary, Indonesia, Ireland, Italy, Jordan, Kuwait, Latvia, Lithuania, Luxembourg, Malaysia, Mali, Malta, Mauritania, 6Mauritius, Netherlands, New Zealand, Niger, Nigeria, Norway, Oman, Poland, Portugal, Qatar, Romania, Saudi Arabia, 7Senegal, Singapore, Slovakia, Slovenia, Spain, Sri Lanka, St Lucia, Sweden, Syria, the United Arab Emirates, United 8Kingdom.

2 12.1.3 Releases to the environment

2As a result of the use of endosulfan as an insecticide, endosulfan is released to the environment. No natural 3sources of the compound are known. From the manufacture of formulation operations, local scale 4environmental releases to the air, waste water, or surface waters may also occur.

5 Global usage and emission of endosulfan, and the relationship between global emissions and the air 6concentration of endosulfan in the Canadian Arctic were reported in Li and MacDonald (2005). Cumulative 7global use of endosulfan for crops is estimated to be 338 kt (103 tonnes). The average annual endosulfan 8usage in the world is estimated to have been 10.5 kt from 1980 to 1989 and 12.8 kt from 1990 to 1999. The 9general trend of total global endosulfan use has increased continuously since the first year this pesticide was 10applied. No recent figures, updated after the recent banning in several countries, are available. India is the 11world's largest consumer of endosulfan with a total use of 113 kt from 1958 to 2000. Total global endosulfan 12emissions have also increased continuously since the year when this pesticide was first applied presently 13amounting to an estimated total emission around 150 kt.

14A time trend of α-endosulfan air concentration at Alert between 1987 and 1997 (Li and MacDonald (2005)), 15compiled from several sources (Patton et al., 1989, Halsall et al., 1998 and Hung et al., 2002), shows this to 16be one of the few organochlorine pesticides that is still increasing in Arctic air. The data for emissions of α- 17endosulfan exhibit high variability but demonstrate a generally increasing trend at least up until the late 181990s. Canadian Arctic air sampling data similarly exhibits high variability but the few available data are not 19inconsistent with the emission data, suggesting the atmosphere is an important transporting medium.

20 2.2 Environmental fate

212.2.1 Persistence

22Endosulfan aerobic transformation occurs via biologically mediated oxidation. The main metabolite formed is 23endosulfan sulfate. This compound is slowly degraded to the more polar metabolites endosulfan diol, 24endosulfan lactone, endosulfan ether. Formation of endosulfan sulfate is mediated essentially by micro- 25organisms, while endosulfan-diol was found to be the major hydrolysis product. Microbial mineralisation to 26carbon dioxide is generally slow.

27Endosulfan sulfate also has insecticidal activity. Given the comparable toxicity of the sulfate metabolite a 28number of authors make use of the term “endosulfan(sum)” which includes the combined residues of both 29parent isomers and endosulfan sulfate. However, this term does not consider that in reality all the metabolites 30of endosulfan retain the backbone of the structure with the hexachloronorbornene bicycle.

31The following degradation patterns for soil (right figure) and water (left figure) are proposed in the European 32Union risk assessment. In both cases, the parent isomers are transformed in endosulfan diol, either directly or 33through endosulfan sulfate. Endosulfan diol is then degradated into a set of related metabolites, including 34endosulfan ether, endosulfan hydroxyether, endosulfan carboxylic acid, and endosulfan lactone.

O Cl S O Cl Cl O Cl a- and ß- endosulfan

O CH O H C OH 2 2 Cl O Cl S O Cl Cl Cl Cl Cl Cl O Cl Cl CH Cl Cl 2 CH2OH Cl Cl Cl Cl Cl Cl

endosulfan sulfate endosulfan diol endosulfan ether

O OH H COOH Cl Cl O Cl Cl Cl Cl O Cl Cl Cl Cl Cl CH 2 Cl CH2 CH 2OH Cl Cl Cl Cl Cl Cl

hydroxy endosulfan endosulfan lactone endosulfan carboxylic acid hydroxyether

CO2 + Unknown metabolite(s) + Bound residue 1 2 3This environmental fate complicates the assessment of persistence as DT 50 values. Most studies suggest that 4α-endosulfan has a faster degradation than β-endosulfan, and that endosulfan sulfate is much more persistent.

5There is high variability in the reported DT50 values for these substances. In the European Union assessment, 6the reported DT50 for aerobic soil degradation under laboratory conditions, ranged from 25 to 128 days for the 7α + β isomers, and from 123 to 391 days for endosulfan sulfate. The rapid field dissipation of endosulfan 8following its application under normal conditions is mostly related to volatilisation and varies largely; the

2 1European Union assessment reported, for the temperate regions, field DT50s ranging from 7.4 to 92 days for 2the α + β isomers. A fast dissipation has been observed for tropical climates; volatilization, particularly for the 3α and β isomers, is considered the major process for endosulfan dissipation in tropical environments (Ciglasch 4et al., 2006; Chowdhury et al., 2007). Field aging increases the persistence in soil and is particularly relevant

5for endosulfan, with a 3-fold increase in the apparent organic carbon partition coefficient K OC within 84 days 6in a tropical fruit orchard under natural weather conditions (Ciglasch et al., 2008).

7At POPRC 4, the combined DT50 measured in laboratory studies for α and β endosulfan and endosulfan 8sulfate, was selected as a relevant parameter for quantifying the persistence of endosulfan. A large variability 9on the rate of this degradation has been observed. The estimated combined half-life in soil for endosulfan (α, 10β isomers and endosulfan sulfate) ranges typically between 28 and 391 days; but higher and lower values are 11reported in the literature under specific conditions.

12In the aquatic compartment, endosulfan is stable to photolysis. A rapid hydrolysis is only observed at high pH 13values, and it is non-readily biodegradable. In water/sediment systems (Jones, 2002; 2003 reported in the EU 14dossier) DT50s for the alpha, beta isomers and endosulfan sulfate ranging between 3.3 and 273 days, were 15presented. These specific values were not validated but DT50 > 120 d was demonstrated. Endosulfan diol, and 16under acidic conditions endosulfan lactone, were also observed at relevant levels.

17There is a high uncertainty regarding the degradation rate of endosulfan in the atmosphere. Buerkle (2003) 18has presented a set of estimations based on Structure Activity Relationship and experimental values. A half 19life estimation in atmosphere using the Atkinson method was conducted in 1991, resulting in a value of 8.5 d 20but with high uncertainty. Experimental figures are presented for α-endosulfan (27 d at 75ºC for flashlight 21photolysis) and β-endosulfan (15 d based on the Freon-113 method). The AOPWIN calculation method 22indicates a half life of 47.1 hours assuming a constant diurnal OH concentration of 5 x105 cm-3.

23It is concluded that, considering endosulfan and its related transformation products, the persistence of 24endosulfan in soil, sediments and air is confirmed.

252.2.2 Bioaccumulation

26Three complementary sources of information have been analysed for assessing the bioaccumulation and 27biomagnification potential of endosulfan and its degradation products: the screening assessment based on 28physical-chemical properties, the analysis of experimental data, including bioconcentration, bioaccumulation 29and toxicokinetic studies, and the analysis of field collected information. The key elements of these 30assessments are presented below.

31Screening assessment based of physical-chemical properties

32The reported log Kow for α- and β-isomers and endosulfan sulfate range between 3 and 4.8. New studies 33(Muehlberger and Lemke 2004) using the HPLC-method indicates a log K ow of 4.65 for α-endosulfan, 4.34 34for β-endosulfan and 3.77 for endosulfan sulfate. The other metabolites included in the Kow determination 35have lower Kow than endosulfan sulfate. These values indicate potential for bioconcentration in aquatic 36organisms, although they are below the screening trigger of 5 for the Stockholm Convention.

37Recently, the role of the octanol/air partition coefficient (Koa) for the screening assessment of the 38biomagnification potential of POPs in terrestrial food chains is receiving significant attention. Kelly & Gobas 39(2003) and Kelly et al. (2007) have proposed that the biomagnification of endosulfan in the terrestrial food 40chain is particularly relevant, because it has a high log Koa. A high Koa causes slow respiratory elimination. 41The proposed log Koa for α- and β-endosulfan is 10.29; and for endosulfan sulfate is 5.18. Although there are 42no specific screening thresholds for the Koa the authors suggests that chemicals with a log Kow higher than 2 43and a log Koa higher than 6 have an inherent biomagnification potential in air-breathing organisms of 44terrestrial, marine mammalian, and human food chains. Endosulfan clearly falls within this category.

45Bioconcentration studies in aquatic organisms

2 1The reported BCF values for fish ranged from approximately 20 to 11,600 (L kg-1 wet wt.); however, the 211,600 value (Johnson and Toledo, 1993) is considered of low reliability. A BCF of 5,670 has been proposed 3from the US-EPA re-evaluation of this study, but the uncertainty is still too high and the data should not be 4considered as reliable. The USEPA re-evaluated in 2007 the bioconcentration studies (U.S.EPA 2007). The 5two highest quality studies, based on US Federal Insecticide Fungicide and Rotendicide Act (FIFRA) criteria, 6indicate that the BCF range for fish is 1,000 (striped mullet; Schimmel et al. 1977) to 3,000 (sheepshead 7minnow; Hansen and Cripe 1991). Depuration half-lives in fish for α- and β-endosulfan and endosulfan 8sulfate were 2–6 days. Bioconcentration studies were available for five species of invertebrates in which BCF 9ranged from 12 to 600. An average BCF of 2,682 and 3,278 (dry weight) was determined for freshwater green 10algae and Daphnia magna, respectively (DeLorenzo et al. 2002). It should be noted that D. magna neonates 11accumulated little endosulfan when exposed via the ingestion of contaminated phytoplankton.

12The assessment of parent and metabolite bioconcentration is particularly relevant. The study by Pennington et 13al., (2004) offers a good example of the complexity of these estimations. Oysters were exposed to endosulfan 14in an estuarine mesocosm for 96h. Within this short exposure period, a significant bioaccumulation of α- and 15β-endosulfan in oysters is observed, but the quantification, even under mesocosm controlled conditions, is 16very different depending on how the water and organisms concentrations are compared. The authors suggest 17BCF values between 375 and 1776 (dry weight) for total (α-, β- and endosulfan sulfate). An outdoor aquatic 18microcosms study has been presented in the CropLife dossier (Schanne, 2002). The study was conducted 19outdoors in order to simulate the conditions in natural systems as closely as possible. For that purpose, 20sediment, water and other biota were collected from a large, shallow water natural reserve area of the Austrian 21side of Lake Constance. Concentrations of radio-labelled endosulfan lactone, and two unknown metabolites, 22M1 and M4, in water increased constantly during the study, whereas endosulfan sulfate was more or less 23constant at a low level or slightly decreasing at both entry routes. The total radioactive sediment residue was 24increasing during the study to maximum 13.8 µg radioactivity equivalents/kg. The total radioactive residue in 25macrophytes increased with time reaching a maximum of 2236 µg radioactivity equivalents kg -1 fresh weight. 26Like for macrophytes, the total radioactive residue in surviving fish reached a maximum of 3960 µg 27radioactivity equivalents kg-1 fresh weight.

28This study clearly demonstrates that endosulfan is found in the sediment, fish and macrophytes up to study 29termination and is also degraded to metabolites that maintain the chlorinated cyclic structure of endosulfan. 30These metabolites have the potential to bioaccumulate in fish and macrophytes, and some of them have 31demonstrated their potential for persistence in the environment. In addition to this, the study reveals that there 32are other unknown metabolites with the same potential for bioaccumulation. The bioaccumulation factors 33(BAF) for spray-drift and run-off routes were estimated as: BAF total radioactivity ca.1000; BAF endosulfan- 34sulfate 4600-5000 (spray-drift). It should be noted that these BAFs should be taken with care as the tested 35concentrations provoked clear effects on aquatic organisms or were too close to toxic concentrations; 36therefore, the estimated bioaccumulation potential could be different to that expected due to the toxic effects 37of the tested concentrations.

38Toxicokinetic and metabolism studies

39Following oral administration of endosulfan, either via single oral dose or dietary administration, elimination 40of the parent compound and its metabolites is extensive and relatively rapid in a range of species of 41experimental animals. The metabolites of endosulfan include endosulfan sulfate, diol, hydroxy-ether, ether, 42and lactone.

43A physiologically based pharmacokinetic model for endosulfan metabolism in the male Sprague-Dawley rats 44has been developed by Chan et al. (2006). Recently, the accumulation and elimination kinetics of dietary 45endosulfan in Atlantic salmon has been published (Berntssen et al., 2008). Dietary β-endosulfan showed a 46higher biomagnification factor (BMF) (0.10±0.026 vs. 0.05±0.003, p<0.05) than α-endosulfan, with a higher 47uptake (41±8% vs. 21±2%) and lower elimination (26±2 x 10-3 day-1 vs. 40±1 x 10-3 day-1) rate constants, and. 48Endosulfan sulphate levels remained unchanged during the depuration period, whereas the parental isoforms 49were rapidly eliminated. Based on the decrease in diastereomeric factor over time, biotransformation was 50estimated to account for at least 50% of the endosulfan elimination. The formation of the metabolite

2 1endosulfan sulfate comprised a maximum 1.2% of the total accumulation of endosulfan. No other metabolites 2were measured, and therefore a BMF for endosulfan plus all metabolites cannot be estimated form this study.

3Assessment of field data and biomagnification models

4A large number of studies offering information on measured levels of endosulfan in biota all over the world 5are available. Endosulfan and its metabolite endosulfan sulfate are frequently found in crops and in the 6vicinity of treated sites, as well as in remote areas where the presence of this pesticide must result from 7medium and long range transport from those areas in which endosulfan has been used.

8Quantitative estimates of biomagnification can be obtained through the use of mathematical models calibrated 9with field data (Alonso et al., 2008). Several published models indicate the potential biomagnification of 10endosulfan through the food chain. A model of the lichen-caribou-wolf food chain predicts biomagnification 11of β-endosulfan. The BMFs for wolf range from 5.3 to 39.8 for 1.5 to 13.1 year old wolves (Kelly et al. 2003).

12A particularly relevant piece of information was published in 2007 (Kelly et al., 2007). The model predicts a 13significant BMF for β-endosulfan in air-breathing species, ranging from 2.5 for terrestrial herbivores to 28 for 14terrestrial carnivores; and BMF below 1 for water-respiring organisms.

15Also in the Canadian Arctic concentrations of α-, and β-endosulfan and endosulfan sulfate in ice-algae, 16phytoplankton, zooplankton, marine fish and ringed seal have been presented. Concentrations ranged from 170.1 – 2.5 ng g-1 lipid. Calculated trophic magnification factors were less than 1, suggesting no 18biomagnification in the ringed seal food chain (Morris et al. 2008). However a trophic magnification factor >1 19was calculated for the Southern Beaufort Sea and Amundsen Gulf food webs if marine mammals are included 20in the food web (Mackay & Arnold (2005).

21The comparison of reported concentrations of endosulfan in biota, and particularly in top predators, with those 22observed in the same organisms and ecosystems for other POPs, also offer indirect indications of 23bioaccumulation potential. Although the numerical BCF threshold is not exceeded in the standard laboratory 24studies, there is enough information demonstrating that the bioaccumulation potential of endosulfan is of 25concern, particularly for terrestrial food chains.

262.2.3 Potential for long-range environmental transport

27The potential of endosulfan for long range transport can be evaluated from three main information sources, 28the analysis of the endosulfan properties, the application of LRT models, and the review of existing 29monitoring data in remote areas.

30Screening of physical-chemical properties

31There is enough information on the volatility of α and β endosulfan to support the potential for atmospheric 32transport. Atmospheric transport over long distances requires a minimum level of persistence in the 33atmosphere; as presented above, there is uncertainty on the real degradation rate of endosulfan in the 34atmosphere but the threshold half life of 2 days seems to be exceeded. Taking into account the much lower 35temperatures of the troposphere, the environmental half-life of endosulfan under real situations is likely to be 36even longer. Therefore, it should be concluded that the combination of volatility and sufficient atmospheric 37persistence results in a significant potential for long range transport.

38LRT model predictions

39Several models have been developed for estimating this potential according to the characteristics of the POP 40candidates. Becker, Schenker and Scheringer (ETH, 2009 Swiss submitted information) have estimated the 41overall persistence (POV) and LRT potential (LRTP) of α- and β-endosulfan and two of their transformation 42products, endosulfan sulfate and endosulfan diol with two multimedia box models, the OECD POV and LRTP 43Screening Tool and the global, latitudinally resolved model CliMoChem. The OECD Tool yields POV and

2 1LRTP for each compound separately, whereas the CliMoChem model calculates the environmental 2distribution of the parent compounds and the formation and distribution of the transformation products 3simultaneously. Results from the CliMoChem model show that POV and LRTP of the endosulfan substance 4family are similar to those of acknowledged POPs, such as aldrin, DDT, and heptachlor. The results also show 5that POV and LRTP of the entire substance family, i.e. including the transformation products, are 6significantly higher than those of the parent isomers alone.

7The US (US submitted information) concludes that recent studies suggest that desorbed residues of 8endosulfan volatilize and continue to recycle in the global system through a process of migration and re- 9deposited via wet and dry depositions as well as air-water exchange in the northern hemisphere. Dust 10dispersion and translocation also contribute endosulfan into the atmosphere as adsorbed phase onto suspended 11particulate matter, but this process does not appear to be a major contributor like volatilization. Transport of 12endosulfan in solution and sediment bound residues also can potentially contribute in the long-range and 13regional distributions of endosulfan.

14Brown and Wania (2008) have recently published model estimations for the Arctic; according to the model, 15endosulfan was found to have high Arctic contamination and bioaccumulation potential and matched the 16structural profile for known Arctic contaminants. These results are in agreement with the empirical 17estimations of Arctic contamination potential reviewed by Muir et al (2004) which concluded that endosulfan 18is subject to LRT as predicted by models and confirmed by environmental measurements.

19Confirmation based on measures in remote areas

20This potential has been confirmed by monitoring data; there is a significant amount of information as 21endosulfan has been measured in combination with other organochlorine insecticides. Several publications 22indicate the potential for long-range transport of endosulfan residues, and report findings of endosulfan in the 23Arctic at increasing levels in water, air and biota.

24 2.3 Exposure

252.3.1 Environmental monitoring data

26Although endosulfan has only recently been included in formal POP monitoring programs, the chemical is 27frequently measured in studies on organochlorine pesticides, and therefore there is abundant but highly 28variable information on measured levels of endosulfan in environmental samples. Most studies include α- and 29β-endosulfan, and in some cases, endosulfan sulfate is also measured. Other endosulfan metabolites are only 30rarely quantified. The information has been compiled in three main categories:

31 . Medium range transport: Collects the information in untreated areas in the vicinity of areas for which 32 endosulfan has been used or has been potentially used (areas with intensive agricultural activity).

33 . Potential for long range transport: Collects information in areas at significant distance from use 34 areas, where the presence of endosulfan can only be explained by atmospheric transfer and 35 deposition; includes high altitude mountain areas.

36 . Long range transport: Collects information in remote areas, far away from intensive use areas, in 37 particular, the Arctic and the Antarctic.

38A summary of relevant monitoring values is presented below. This summary is mostly based on the recent 39reviews by the European Communities and the USA submitted within their information dossiers, and 40completed by additional information presented by other parties/observers and the review of recent literature 41data.

42Medium term transport: water and aquatic organisms

2 1Since 1991, in the USA, the South Florida Water Management District’s (SFWMD) non-targeted quarterly 2water quality monitoring program has been analyzing a number of pesticides including endosulfan at 34 sites. 3Endosulfan and endosulfan sulfate were detected in surface waters and benthic sediments at several locations 4in the south Miami-Dade County farming area. Endosulfan has been measured at concentrations exceeding 5Florida’s chronic surface water quality standard of 0.056 μg L-1 for a number of years.

6In 1997, pesticide concentrations were measured in mountain yellow-legged frogs (Rana muscosa) from two 7areas in the Sierra Nevada Mountains of California, USA. From LeNoir et al (1999) these results support the 8hypothesis that contaminants have played a significant role in the decline of R. muscosa in the Tablelands of 9Sequoia National Park. The University of South Carolina (USC) and the National Oceanic and Atmospheric 10Administration (NOAA) also conducted a monitoring study targeting areas where endosulfan was used 11(Delorenzo et al., 2001). The data indicate that total endosulfan residues have moved to areas distant from 12where it was initially applied and that the residues are sufficiently high, when compared to toxicity values of 13aquatic organisms to exceed the Office of Pesticide Programs’ (OPP) acute and chronic risk levels of concern.

14Water samples from four temperate lakes in south-central Canada show the presence of α-and β-endosulfan 15(Muir et al., 2004). Mean concentration levels of α-endosulfan ranged from 1.3 to 28.5pg L -1, and those of β- 16endosulfan from 0.0 to 10.3 pg L-1 in lakes Opeongo, Nipigon, Britt Brook, and Virgin Pond. No agricultural 17area was within 31 miles (50 Km) of any of these lakes, suggesting that the presence of endosulfan resulted 18from atmospheric transportation and deposition. Monitoring and modelling results suggest that under the 19weather conditions prevailing in south-central Canada, endosulfan can potentially undergo regional-scale 20atmospheric transport and reach lakes outside endosulfan use areas.

21Medium term transport: Air and airborne particles

22Detailed atmospheric concentrations of α-endosulfan and β-endosulfan were summarized by Ngabe and 23Bidleman (2001) in North America. Early measurements of endosulfan in air were made during a survey of 24airborne pesticides across the United States in 1970 (Majewski and Capel, 1995). Mean concentrations of α- 25endosulfan ranged from 0.7 ng m-3 in Meadow, North Carolina, to 159 ng m-3 in Peaksmill, Kentucky. The 26average concentrations of α- and β-endosulfan in air were 0.170 and 0.045 ng m-3, respectively, at Solomons, 27Island, Maryland, in 1995 (Harman-Fetcho et al., 2000). The frequency of occurrence of α- and β-endosulfan 28in monitoring samples was 100%.

29Air Resource Board (ARB) of California monitored edge of the field drift of endosulfan just after application 30to an apple orchard in San Joaquin County in April 1997. In a separate study, ambient air was monitored 31during a period of high use of endosulfan in Fresno County in July-August 1996. Air concentrations of α- 32endosulfan ranged from 3800 ng m-3 to 290 ng m-3 adjacent to the treated field. The detections for β- 33endosulfan during the same sampling period ranged from 200 ng m-3 to 48 ng m-3. The ratio of α-isomer: β- 34isomer varied from 5 to 209 across all the samples with concentrations of both isomers above the limit of 35quantification (LOQ).

36Abundant regional air data are available for the Great Lakes Region from a joint US EPA / Environment 37Canada-monitoring project IADN (Integrated Atmospheric Deposition Network) (Sun et al., 2006) and Sun et 38al. (2003) providing compelling evidence for medium-range airborne transport of endosulfan and endosulfan 39sulfate. The endosulfan concentrations (shown as the sum of α- and β-endosulfan) in vapour phase showed a 40clear increasing trend consistent with prevailing winds from the west to east, except for the remote site of 41Burnt Island. At each site, the average concentration was skewed by high outliers that usually occurred in the 42summer and were attributed to current agricultural use of endosulfan. Higher endosulfan concentrations were 43observed at Point Petre, Sturgeon Point, and Sleeping Bear in vapour, particle, and precipitation phases, 44which could be explained by its heavy usage in the surrounding areas (Hoh and Hites, 2004). For example, 45endosulfan is widely used in Michigan and New York State (Hafner and Hites, 2003) and in Ontario (Harris, 46et al., 2001), particularly in the southern and western portions of the province. Total endosulfan 47concentrations in the particle phase declined at all five U.S. sites. Because of the lack of updated usage data, 48correlation between the decreasing particle-bound endosulfan concentrations and its usage pattern is difficult.

2 1However, it should be noted that total endosulfan concentrations showed no long-term decreasing trends in 2the vapour phase at Eagle Harbor, Sleeping Bear Dunes, or Sturgeon Point.

3Shen et al. (2005) evaluated endosulfan concentration in air using passive air samplers (PAS) to trap 4endosulfan. Gaseous concentrations of endosulfan varied from 3.1 to 681 pg m-3 for α-endosulfan and from 50.03 to119 pg m-3 of β-endosulfan. The maximum measured concentration of endosulfan in air was generally 6lower than 58 pg m-3 across North America. The highest measured concentrations were reported in the 7Okanagan Valley, British Columbia, East Point on Prince Edward Island, Manitoba, and Tapachula, Mexico.

8Total endosulfan concentrations also showed a strong seasonal variation in precipitation. The ratio between 9the highest and the lowest total endosulfan concentration ranged between about 2-10. In particular, this ratio 10is as high as 10 at Point Petre, suggesting a heavy usage in the surrounding area. At all sites, the total 11endosulfan concentrations peaked in early July in precipitation, a time which corresponds well with its 12maximum agricultural usage.

13Medium term transport: Rainwater and snow

14Several studies demonstrated that endosulfan is removed from the atmosphere by rain and snow fall. In a 15monitoring study carried out in eastern Canada between 1980 and 1989, α-endosulfan was occasionally 16reported at concentrations near the detection limit of 10 ng L-1 (Brun et al. (1991). In precipitation of the Great 17Lakes region, α- and β-endosulfan concentrations were regularly determined by IADN at various stations 18during the period of 1987–1997 (USEPA, 2007). Concentration levels of α-endosulfan ranged from 0.13 – 191.95 ng L-1 and those of β-endosulfan from 0.19 – 6.09 ng L-1 in Lake Superior and Lake Erie. Higher values 20were reported from Lake Michigan ranging from 0.54 – 8. ng L-1 22 ng L-1 for α- and from 1.06 – 12.13 ng L-1 21for β-endosulfan. Concentrations of the transformation product endosulfan sulfate measured in precipitation 22of the Great Lakes region were mostly in a range of 0.1 to 1 ng L-1.

23Endosulfan and endosulfan sulfate were detected in seasonal snowpack samples at six national parks in the 24Western United States (Hageman et al., 2006). Concentrations of total endosulfan were were measured from 25all sites and ranged from < 0.0040 ng L-1 to 1.5 ng L-1 in the Sequoia, Mount Rainier, Denali, Noatak-Gates, 26Glacier and Rocky Mountain National Parks. The percentage contribution of endosulfan sulfate to the total 27endosulfan concentration ranged from 4.0% to 57.0% with mean value being 24.0%. The study results suggest 28that current use of endosulfan plays a significant role in contributing to the deposition of endosulfan via snow 29to remote high-elevation and high-latitude ecosystems.

30Medium term transport: Sediment

31The presence of endosulfan in benthic sediment is well documented in the National Sediment Contaminant 32Point Source Inventory (NSI) databases prepared by the Office of Science and Technology (OST) of US EPA 33(EPA-823-C-01-001) (Cited in USEPA, 2007). EPA’s evaluation of the NSI data was the most geographically 34extensive investigation of sediment contamination ever performed in the United States. In the NSI data base, 35199 detections for α-endosulfan, ranged from

39 Potential for Long-range transport: Mountainous Regions

40The effect of "global distillation" is believed to account for transport of POPs whereby a compound 41volatilizes from warmer regions, undergo long-range atmospheric transport, and subsequently re-condense to 42an accumulation of these substances in the temperate, higher mountainous and Arctic regions. Wania and 43Mackay (1993) suggested that, through “global distillation” organic compounds could become latitudinally 44fractionated, “condensing” at different temperatures according to their volatility, so that compounds with 45relative low vapour pressures might accumulate preferentially in polar regions. Endosulfan was found in the 46atmosphere of European mountain areas (Central Pyrenees and High Tathras). Like hexachlorocyclohexane

2 1(HCH), endosulfan was found in higher concentrations in the warm periods (4-10 pg m-3) in both the gasand 2particulate phase, reflecting its seasonal use pattern (van Drooge et al. 2004). Many POP substances as well as 3endosulfan were found in snowpack samples collected at different altitudes of mountains in western Canada. 4The levels of contaminants in snow and in snowpack increased with the altitude, showing a 60-100 fold 5increase in net deposition rates of contaminants to snowpack over a 2300 meter rise in elevation (Blais et al., 61998). The concentration range of α-endosulfan was 0.06–0.5 ng L-1 in the sampling altitude range of 700 – 73,100 m. Aerial transport also caused contamination of snow (Sequoia National Park) and water (Lake Tahoe 8basin) of the Sierra Nevada Mountains in California, a region adjacent to California’s Central Valley which is 9among the heaviest pesticide use areas in the U.S. Levels of α-endosulfan found in rain were in a range of < 100.0035 ng L-1 to 6.5 ng L-1 while β-endosulfan was measured at concentrations of < 0.012 ng L-1 up to 1.4 ng 11L-1 McConnell et al. (1998). Concentrations of 71.1 pg m-3 of α-endosulfan were measured in the Himalayas; 12backward trajectory analysis indicted that it arrived there from the Indian subcontinent on westerly winds, 13driven by the Asian monsoon (Li et al 2007).

14For mountain lakes in the Alps, Pyrenees (Estany Redò) and Caledonian Mountains (Øvre Neådalsvatn 15(Norway)), atmospheric deposition of endosulfan was estimated between 0.2 and 340 ng m -2 per month 16(Carrera et al., 2002). Unlike other chemicals, endosulfan showed a more uniform geographical distribution, 17the lakes in the South were much more exposed to endosulfan impact, reflecting the impact of agricultural 18activities in southern Europe. In the northern lakes only the more persistent endosulfan sulfate was measured.

19 Long-range transport: Arctic and Antarctic Areas

20The US review summarizes information by GFEA (2007); Ngabe and Bidleman 2001, and Endosulfan Task 21Force (ETF) report MRID 467343-01.

22Long range atmospheric transport of α- and β-endosulfan to the Arctic was first reportedd in 1986–(Patton et 23al. 1989). A “brown snow” event occurred in the central Canadian Arctic during the year 1988. The snow was 24coloured by dust that appeared to be transported from western China. Endosulfan was detected in the dust at a 25maximum concentration of 22 pg L-1. Since then endosulfan has been routinely found in the Canadian Arctic 26air monitoring program, from 1993 up to the present (Halsall et al., 1998; Hung et al., 2001). Extensive 27monitoring data on endosulfan from the Arctic are available for the atmosphere, snowpack, surface water and 28biota (Bidleman et al., 1992; De Wit et al., 2002; Halsall et al., 1998; Hobbs et al, 2003; Jantunen and 29Bidleman, 1998).

30 Long-range transport: Arctic Air

31Endosulfan was reported as a widely distributed pesticide in the atmosphere of northern polar region. Unlike 32for most other organochlorine pesticides that have declined average concentrations of endosulfan in the Arctic 33have not changed significantly during the last five years prior to the publication (Meaking, 2000). 34Concentrations of α-endosulfan from Arctic air monitoring stations increased from early to mid-1993 and 35remained at roughly at 0.0042-0.0047 ng m-3 through the end of 1997. No clear temporal trends of endosulfan 36concentrations in the arctic atmosphere were observed (Hung et al., 2002). Measurements taken in air at Alert 37in, Nunavut, Canada resulted in annual average concentrations between 3 and 6 pg m -3 during 1993 to 1997. 38Fluctuating values mirror the seasonal applications in source regions.

39Concentrations of endosulfan in Arctic air were found to be exceeded only by those of ΣHCH-isomers and 40hexachlorobenzen (HCB) (Halsall et al., 1998). In comparison to monitored concentrations in the Great Lakes 41region, atmospheric levels in the Arctic were less dependent on temperature, although seasonal variations 42were apparent as well. For example α-endosulfan concentrations ranged by a factor of 3-5 over spring to fall 43periods. This infers a more blurred bimodal seasonal cycle with growing distance from areas of application. 44Hung et al. (2002) used temperature normalization, multiple linear regression, and digital filtration to analyze 45the temporal trends of an atmospheric dataset on organochlorine pesticides collected at the Canadian high 46arctic site of Alert, Nunavut. While air concentrations of lindane and chlordane showed decreasing trends 47through the 1990s with half-lives of 5.6 and 4.8 years α- endosulfan showed a very slow decline with a half- 48life of 21 years.

2 1Seasonal variation of concentrations was also reported from Sable Island (240 km east of Nova Scotia, 2Canada, at 43°57´N, 60°00´W). In summer, aerial endosulfan concentrations (α- and β-isomer) were 3determined between 69 and 159 ng m-3 while for wintertime values dropped to 1.4-3.0 pg m-3 (only α –isomer) 4(Bildeman et al., 1992).

5Similar data on α-endosulfan have been reported from Resolute Bay (Cornwallis Island, 75 N lat.) where air 6concentrations of approximately 4 pg m-3 have been measured (Bidleman et al., 1995) and from air samples 7taken on an iceberg that calved off the Ward Hunt Ice Shelf on the northern shore of Ellesmere Island, 8Canada, (approx. 81°N, 100°W). Mean concentration of α-endosulfan in summer 1986 and 1987 were 7.1 and 93.4 ng m-3, respectively (Patton et al., 1989). Additional evidence for airborne long-range transport is 10provided by data from Newfoundland showing mean concentrations of 20 pg m-3 in summer 1977 (Bidleman 11et al., 1981).

12Further air concentrations of endosulfan were reported from Amerma (eastern Arctic region of Russia) 13between 1–10 pg m-3 (De Wit et al., 2002; Konoplev et al., 2002). Endosulfan was detected in around 90% of 14all samples displaying a significant correlation with atmospheric temperature. Unlike other organochlorines 15where seasonal enhancements are hypothesized to be due to (re)volatisation from secondary sources, fresh 16applications were assumed to be responsible for endosulfan concentrations of 3.6 pg m-3 in winter and 5.8 pg 17m-3 in summer (mean values). Spatially, the annual concentrations at the various circumpolar sites did not 18show remarkable differences, indicating a degree of uniformity in contamination of the Arctic atmosphere.

19 Long-range transport: Arctic Freshwater

20Endosulfan (isomer unspecified) was measured also at Amituk Lake on Cornwallis Island, NV, Canada. The 21ranges were (in ng L-1) 0.135 – 0.466 in 1992, 0.095 – 0.734 in 1993, and 0.217 – 0.605 in 1994 (quoted in 22Ngabè and Bidleman 2001). Annual summertime peaks in endosulfan concentrations observed were attributed 23to fresh input from snow smelt via influent streams.

24 Long-range transport: Arctic Freshwater Sediment

25Laminated cores collected from Arctic Lake DV09 on Devon Island, in Nunavut, Canada, in May 1999 were 26analysed inter alia for endosulfan. Only α-endosulfan was present in the sediment of that lake. The 27concentration was highest at the sediment surface, and rapidly decreased to below detection limits in core 28slices dated prior to 1988.

29 Long-range transport: Arctic Seawater

30Endosulfan was repeatedly detected in Arctic seawater during the 1990s. Mean concentrations were similar 31to those of chlordane and ranged from 2-10 pg L-1. Seasonal trends displayed increasing concentrations during 32the open water season suggesting fresh input from gas exchange and runoff. This trend parallels seasonal 33trends observed in Arctic air and Amituk Lake.

34A survey of several pesticides in air, ice, fog, sea water and surface micro-layer in the Bering and Chuckchi 35Seas in summer of 1993 (Chernyak et al., 1996) identified α-endosulfan in air and subsurface seawater at 36levels around 2 pg L-1. In melted ice less than 9 pg L-1 and for the sea water surface micro-layer less than 40 37pg L-1 were detected. For fog condensates from several sites of that region concentration of <10 to <0.5 ng L-1 38were reported. β-endosulfan was found in several atmospheric samples, e.g. from the Central Bering or Gulf 39of Anadyr at concentrations around 1 pg m-3. Similar concentrations of endosulfan have been reported from 40seawater surface layer (40-60 m) collected in the Bering and Chukchi Seas, north of Spitzbergen and the 41Greenland Sea (Jantunen and Bidleman, 1998).

42Arctic seawater concentrations of endosulfan were measured from 1990s to 2000 in different regions of the 43Arctic Ocean (Weber et al., 2006). Surface seawater concentrations for α- and β-endosulfan ranged from <0.1 44to 8.8 pg L-1 and 0.1 to 7.8 pg L-1 respectively. Geographical distribution for α-endosulfan revealed that the 45highest concentrations in the western Arctic, specifically in Bering and Chukchi Seas with lowest levels

2 1towards the central Arctic Ocean. The results of air-water fugacity ratio indicate that α-endosulfan has been 2undergoing net deposition to surface waters across all the regions of the Arctic Ocean since 1990s. The 3authors concluded that the net deposition through air-water transfer may be the dominant pathway into the 4Arctic Ocean for α-endosulfan, particularly during the ice free periods.

5 Long-range transport: Arctic Snow and Snowpack

6Concentrations of α-endosulfan in snow samples collected in the Agassiz Ice Cap, Ellesmere Island, Canada 7in 1986 and 1987 were measured at concentrations ranging from 0.10 to 1.34 ng m -3 (Gregor and Gummer, 81989). The concentrations of α- endosulfan in snowpack in Agassiz Ice Cap were 0.288 ng L -1 in 1989 and 90.046 ng L-1 in 1992 (Franz et al., 1997). From measured snowpack concentrations and snowfall amounts, 10minimum winter deposition rates of 0.03 μg m-2 were estimated for the years 1986 and 1987 (Barrie et al., 111992).

12 Long-range transport: Arctic and Antarctic Biota

13α-Endosulfan was found in 40% of samples of Antarctic krill. The geometric mean level detected was 418 pg 14g-1 lw, the maximum was 451 pg g-1 lw (Bengston et al., 2008).

15Endosulfan (α- and β-isomer) was found in many different species in Greenland. The highest median and 16maximum concentrations in ng g-1 lw for various tissues and locations per species are summarized here: 17Terrestrial species: ptarmigan (median 1.9 and max 3.0 in liver), hare (median 0.55 and max 0.64 in liver), 18lamb (median n.d. and max 0.65 in liver), caribou (median 0.17 and max 0.39 in muscle), muskox (median 190.016 and max 1.8 in blubber). In freshwater fish: Arctic char (median 21 and max 92 in muscle tissue). In 20marine organisms: shrimp (median 3 and max 5.2 in muscle), snow crab (median 19 in muscle and max 95 in 21liver), Iceland scallop (median 0.36 and max 1.6 in muscle) capelin (median 50 ng g -1). In seabirds: common 22eider (median 4.9 and max 8.6 in liver), king eider (median 3.7 in liver and max 10 in muscle), kittiwake 23(median 62 and max 130 in muscle), thick-billed murre (median 8.8 and max 15 in liver). In marine 24mammals: ringed seal (median 5.6 in liver at Qeqertarsuaq and max 25 in muscle at Ittoqqortoomiit), harp 25seal (median 12 and max 45 in blubber), minke whale (median 12 and max 29), beluga (median 45 and max 2683 in skin), and narwhal (median 81 and max 120 in skin (Vorkamp et al., 2004).

27Blubber samples from male beluga were collected over 20 years at five time points in Cumberland Sound, 28Canada. Only endosulfan sulfate was detected. But unlike other organochlorines, levels appear to have 29increased steadily (3.2 fold) over that 20 year time period from 1982 reaching ca. 14 ng g -1 lw in 2002. α- 30endosulfan concentrations in blubber of minke whale populations from distinct parts of the North Atlantic 31were sampled in 1998 (Hobbs et al., 2003). The highest mean concentrations were found for minke whales in 32the North Sea/Shetland Islands (34 ng g-1 lipid for females and 43.0 ng g-1 for males), the Barents Sea (7.74 ng 33g-1 lw for females and 9.99 ng g-1 lw for males) and Norway's Vestfjorden/ Lofotes (4.51 ng g-1 lw for females 34and 9.17 ng g-1 lw for males). Lower concentrations of < 1 ng g-1 lw and 5 ng g-1 lw were reported for whales 35from Jan Mayen (territory of Norway) and Greenland. The differences were attributed to distinctions based on 36genetics, fatty acid profiles, etc.

37Endosulfan was detected in adipose tissue and blood of polar bears from Svalbard (Norway). Mean values 38found for α-endosulfan were 3.8±2.2 ng g-1 wet weight (min-max: 1.3-7.8 ng g-1) and 2.9 ± 0.8 ng g-1for β- 39endosulfan (min-max: 2.2-4.3 ng g-1). While the α-isomer was detectable in all samples (15/15) the β-isomer 40was found in just 5 out of 15 samples.

41Alpha-endosulfan ranged between <0.1 and 21 ng g-1 wet weight fat, (<0.1-36 ng g-1lw) in the fat of polar 42bears sampled along the Alaskan Beaufort Sea coast in spring, 2003 (Bentzen et al., 2008).

43In liver of northern fulmar from Bjørnøja endosulfans were detected for just two individuals out of fifteen at 44low levels of 0.28 and 0.50 ng kg-1 lw (Gabrielsen, 2005).

2 1 Levels in murre eggs sampled in 2003 at St. Lazuria Island ranged from 3.04 to 11.2 ng g-1 (mean 5.89 ng g-1) 2for β-Endosulfan while and from 0.116 to 0.428 ng g-1 (mean 0.236 ng g-1) for α-endosulfan. At Middleton 3Island in the Gulf of Alaska, measured levels in 2004 in murre eggs for β-endosulfan ranged up to 11.8 ng g -1 4(mean of 6.74 ng g-1). α- and β-endosulfan were also found in common murre eggs at East Anatuli Island, 5Duck Island, Gull Island, Cape Denbigh, Cape Pierce, Sledge Island, Bluff and Bogoslov Island (Roseneau et 6al., 2008).

7Endosulfan levels in Chinook and sockeye salmon, Cook Inlet Alaska ranged from 252 to 1610 ng kg-1 8(USEPA, 2003).

9In ringed seals from Alaska, the highest levels were found in the western Arctic Ocean off Barrow (geometric 10mean in ringed seal blubber combined males and females of 22.6 ng g-1 α-endosulfan with the upper 11concentration at 43.39 ng g-1) (Mackay and Arnold, 2005).

12Endosulfan has been detected in biota in the Arctic (5 terrestrial, 1 freshwater and 13 marine species with 13maximum levels between 0.39 to 130 pg g-1 lw) and Antarctic (krill with maximum levels of 451 pg g-1 lw). 14Monitoring data have detected endosulfan (and endosulfan sulphate) in the air, the freshwater, the marine 15water and the sediment of the Arctic and/or Antarctic regions. Therefore, there is sufficient evidence that 16endosulfan is transported at long distances and bioccumulates in biota in remote areas.

17 2.4 Hazard assessment for endpoints of concern

18Endosulfan is highly toxic for most invertebrates and vertebrates, including humans. The insecticidal 19properties are shared, with some differences in potency, by the α and β isomers and the metabolite endosulfan 20sulfate. The toxicity of endosulfan has been evaluated by several organizations, including among others 21JMPR in 1998 (FAO/WHO, 1998); ATSDR in 2000 (ATSDR, 2000); the EU in 1999 with addenda up to 222004 (EC dossier submitted as additional information); an EFSA Scientific Panel in 2005 (EFSA, 2005), 23Australia in 2005 (submitted as additional information), Canada in 2007 (PMRA’s REV2007, submitted as 24additional information), US EPA in 2007 (submitted as additional information), and New Zealand in 2008 25(submitted as additional information).

26The toxicity of other endosulfan metabolites has also been demonstrated for different species including 27humans.

28Adverse effects on aquatic organisms

29Endosulfan α, β and sulfate are highly toxic to aquatic invertebrates and fish. Acute median lethal -1 30concentrations (LC50s) for several species at levels below 1µg L have been reported. Chronic no observed 31effect concentrations (NOECs) below 0.1 µg L-1 have been reported for fish and aquatic invertebrates. A 32significant toxicity for aquatic organisms has been also observed for other metabolites; unfortunately, no

33chronic aquatic toxicity data are available for these metabolites, but the acute LC50s for endosulfan lactone and 34ether are lower than 1 mg l-1 (highly toxic to aquatic organisms according to the UN-GHS classification),

35with reported Kow higher than the GHS trigger for chronic classification in the case of endosulfan ether, and 36are not expected to be readily biodegradable..

37The NOEC for sediment dwelling organisms tend to be between 0.1 and 1 mg kg-1, with equivalent pore water 38concentration of about 1µg L-1. The dietary toxicity of endosulfan on fish has been studied in Atlantic salmon 39(Salmo salar), histopathological effects were observed after 35 d of exposure at a diet containing 4 µg kg-1 of 40endosulfan, and the condition factor was significantly reduced in fish exposed for 49 d to 500 µg kg-1 (Petri et 41al., 2006; Glover et al., 2007).

42Additional sublethal effects of particular concern, including genotoxicity and endocrine disrupting effects 43have been reported. Associated genotoxic and embryotoxic effects have been observed in oysters exposed to 44endosulfan (Wessel et al., 2007). Endosulfan sulfate has been shown to be an anti-ecdysteroidal compound for 45Daphnia magna delaying the molting process (Palma et al., 2009). The ecdysteroid system is used by

2 1crustaceans and other arthropods as the major endocrine signalling molecules, regulating processes such as 2moulting and embryonic development. Neurotoxicity has been observed in common toad (Bufo bufo) tadpoles 3(Brunelli et al., 2009), and developmental abnormalities on anuran Bombina orientalis embryos (Kang et al., 42008). In ovum exposure at a critical period for gonadal organogenesis provoked post-hatching effects in 5Caiman latirostris (Stoker et al., 2008). Immunotoxicity has been observed in Nile tilapia (Tellez-Bañuelos et 6al., 2008; Girón-Pérez et al., 2008). Toxic effects have also been observed on non-animal species, including 7cyanobacteria (Kumar et al., 2008) and aquatic macrophytes (Menone et al., 2008).

8Adverse effects on terrestrial organisms

9In laboratory animals, endosulfan produces neurotoxicity effects, which are believed to result from over- 10stimulation of the central nervous system. It can also cause haematological effects and nephrotoxicity. The α- 11isomer was generally found more toxic than the β-isomer (ATSDR, 2000).

12The lowest relevant NOEC for Endosulfan in terrestrial vertebrates is 0.6 mg kg-1 bw day1 based on reduced 13body-weight gain, increased marked progressive glomerulonephritis, blood vessel aneurysm in male rats at 142.9 mg kg-1 bw day-1; the same value was reported in a 1-year dog study. Reproductive effects on mallard 15ducks (Anas platyrhynchos) where observed at low dietary levels, the reported NOEC was 30 ppm in the diet. -1 16The acute median lethal dose (LD50) value in this species is of 28 mg kg bw (see INIA, 1999).

17Toxicity has been shown for bees, beneficial arthropods and soil dwelling invertebrates in the laboratory and 18field studies (i.e., INIA, 1999, New Zealand dossier, Vig et al., 2006; Bostanian and Akalach 2004).

19Adverse effects on human health

20Endosulfan is highly acutely toxic via oral, dermal and inhalation routes of exposure. Exposure through 21certain conditions of use (e.g. lack of protective equipment), and ‘bystander’ exposure have been linked to 22congenital physical disorders, mental retardations and deaths in farm workers and villagers in developing 23countries in Africa, Asia and Latin America. A survey conducted by PAN Africa in Mali in 2001 of villages 24in 21 areas of Kita, Fana and Koutiala found a total 73 cases of pesticide poisoning and endosulfan was the 25main pesticide identified. Endosulfan was found among the most frequently reported intoxication incidents, 26adding unintentionally further evidence to its high toxicity for humans.

27The primary effect of endosulfan, via oral and dermal routes of exposure, is on the central nervous system 28(CNS). Effects in laboratory animals as a result of acute, subchronic, developmental toxicity and chronic 29toxicity studies indicate that endosulfan causes neurotoxic effects, particularly convulsions, which may result 30from over stimulation of the CNS. Possible mechanisms of neurotoxicity include (a) alteration of 31neurotransmitter levels in brain areas by affecting synthesis, degradation, and/or rates of release and reuptake, 32and/or (b) interference with the binding of neurotransmitters to their receptors. Additional effects were noted 33in the liver, kidney, blood vessels and haematological parameters following repeated exposure to endosulfan.

34Acute exposure to high doses of endosulfan results in hyperactivity, muscle tremors, ataxia, and convulsions.

35The LD50 of endosulfan varies widely depending on the route of administration, species, vehicle, and sex of 36the animal. Female rats are clearly more sensitive to endosulfan than male rats, and, on the basis of a single -1 37study, this sex difference appears to apply to mice also. The lowest oral LD50 value is 9.6 mg kg bw in -1 - 38female Sprague-Dawley rats (Ratus norvegicus), and the lowest inhalation LC50 is 0.0126 mg L (2.13 mg kg 391 bw) in female Wistar rats (R. norvegicus). The lowest relevant NOAEL for endosulfan in laboratory animals 40is 0.6 mg kg-1 bw day-1.

41Regarding the metabolites, a particularly relevant study is the 90d toxicity study in rat dietary exposure on 42endosulfan-lactone, conducted by Langrand-Lerche (2003) and included in the EU dossier. The NOEC 43reported in this study is 0.6 mg kg-1 bw day-1, although mild effects in liver and kidney were observed at this 44dose.

2 1Evidence regarding genotoxicity is inconclusive. The assessments conducted by the EU, Canada or the USA 2concluded that endosulfan is not carcinogenic. However, Bajpayee et al., (2006) found that exposure to 3sublethal doses of endosulfan and its metabolites induce DNA damage and mutation. Although the 4contribution of the metabolites to the genotoxicity of the parent compound in bacteria (Salmonella spp.) and 5mammalian cells was unclear, and the pathways leading to bacterial mutation and mammalian cell DNA 6damage appeared to differ.

7Contradictory opinions on the potential for endocrine disruption have been presented. Recent information 8indicates that endosulfan mimics non-uterotrophic E(2) actions, strengthening the hypothesis that endosulfan 9is a widespread xenoestrogen (Varayoud et al., 2008), acts via a membrane version of the estrogen receptor-α 10on pituitary cells and can provoke Ca++ influx via L-type channels, leading to prolactin (PRL) secretion 11(Watson et al., 2007), and is also anti-progestative (Chatterjee et al., 2008).

12It should be noticed that the toxicological reviews have been mostly conducted in the framework of the 13pesticides registrations in various countries. As a consequence, some specific issues, of particular relevance in 14the long-term exposure assessment of POP related characteristics received little attention. For example, in the 15rat chronic study, females from the high dose group had a reduced survival rate after 26 weeks (93% in 16controls, 74% in high dose) and 104 weeks (88% in controls, 46% in high dose). The deaths were 17predominantly associated with respiratory infections. This effect could be associated to the potential 18immunotoxicity of endosulfan that has been hypothesized in some studies. As the study was not designed for 19the specific assessment of these endpoint, relevant effects at low doses could remain unobserved and only 20dramatic effects (over 50% mortality was observed in this case) are evidenced.

21In some chronic toxicity studies, the concentrations of endosulfan and its metabolites were measured at the 22end of the study, but the detection level were too high and only endosulfan sulfate and occasionally 23endosulfan lactone, were above the quantification level. These limitations increase the uncertainty in the 24comparison of measured values in biota with the reported toxicological information.

25There are toxicity and ecotoxicity data available for both endosulfan isomers and several metabolites. 26Endosulfan is a very toxic chemical for many kinds of biota. Metabolism occurs rapidly, but the oxidised 27metabolite endosulfan sulfate shows an acute toxicity similar to that of the parent compound. Endosulfan may 28cause endocrine disruption in both terrestrial and aquatic species. Degradation studies indicate that endosulfan 29is degraded into a large number of other metabolites, all of them retaining the endosulfan structure, and some 30of them showing significant toxicity while others do not. Therefore, there is sufficient evidence that 31endosulfan causes adverse effects to human health and the environment.

323. Synthesis of information

33The potential health and environmental risks of endosulfan associated with its use as a pesticide are well 34documented and have resulted in banning the compound or imposition of severe use restrictions in many 35countries around the world. Human fatality and chronic poisoning cases, and severe environmental problems 36have been reported (Durukan et al., 2009; Jergentz et al., 2004). These potential risks are not limited to the 37crop areas; environmental concentrations representing a potential risk to aquatic species have been found 38associated with medium-range transport of endosulfan. For example, values above the reported NOEC for 39aquatic organisms have been found in Sierra Nevada Mountains of California, USA, (CDPR, 2000). The 40assessment of the POP characteristics of endosulfan confirms the concern regarding endosulfan and its 41metabolites.

42The persistence of endosulfan should be assessed in terms of a dual evaluation. First, the persistence of the 43“active” molecules, with insecticidal activity: the isomers α- endosulfan and β-endosulfan, and the main 44metabolite endosulfan sulfate. Second, the overall persistence of the number of transformation products which 45maintain a similar chemical structure with the hexachloronorbornene bicycle: endosulfan diol, endosulfan 46lactone, endosulfan ether; endosulfan hydroxyether; endosulfan carboxylic acid.

2 1This environmental fate complicates the assessment of persistence using DT50 values. At POPRC 4, the 2combined DT50 measured in laboratory studies for α and β endosulfan and endosulfan sulfate, was selected as 3a relevant parameter. A large variability in the rate of this degradation has been observed. The estimated 4combined half-life in soil for endosulfan (α, β isomers and endosulfan sulfate) ranges typically between 28 5and 391 days; but higher and lower values are reported in the literature under specific conditions. In the field, 6volatilization from soil and plant surfaces is expected to be a main dissipation route

7In the aquatic compartment, endosulfan is stable to photolysis; a rapid hydrolysis is only observed at high pH 8values, and it is non-readily degradable. The dissipation of endosulfan and the abundance of one or other 9degradation products is influenced by the pH and other properties of the water/sediment system. The 10accumulation of endosulfan sulfate in the sediment and of endosulfan hydroxy carboxylic acid in water has

11been seen throughout the studies. The degradation rate could not be estimated, but DT 50 > 120 d has been 12demonstrated. Under acidic conditions endosulfan lactone seems to accumulate in the sediment not reaching 13a plateau after one year. The persistence of endosulfan and other pesticides in aquatic ecosystems of the 14tropics is not substantially lower than during summer in temperate regions.

15There is a high uncertainty on the degradation rate of endosulfan in the atmosphere. However, there is enough 16information on the volatility of α and β endosulfan, and therefore the persistence in the atmosphere is essential 17for supporting the potential for atmospheric transport. The atmospheric transport at long distances requires a 18minimum level of persistence in the atmosphere; despite the uncertainty on the real degradation rate of 19endosulfan in this compartment the threshold half life of 2 days seems to be exceeded. Therefore, it should be 20concluded that the combination of a high volatility and sufficient atmospheric persistence may result in a 21significant potential for long range transport.

22Several models have been developed for estimating this potential according to the characteristics of the POP 23candidate molecules. Results from the CliMoChem model show that POV and LRTP of the endosulfan 24substance family are similar to those of acknowledged Persistent Organic Pollutants, such as aldrin, DDT, and 25heptachlor. The results also show that POV and LRTP of the entire substance family, i.e. including the 26transformation products, are significantly higher than those of the parent compounds alone.

27Several authors have suggested that endosulfan is subject to LRT as predicted by models and posses a high 28arctic contamination and bioaccumulation potential; matching the structural profile for known arctic 29contaminants. The US concludes that desorbed residues of endosulfan volatilize and continue to recycle in the 30global system through a process of migration and are re-deposited via wet and dry depositions as well as air- 31water exchange in the northern hemisphere.

32These suggestions are confirmed by measured data. The presence of endosulfan in remote areas, including the 33Arctic and Antarctic, confirms that endosulfan has enough persistence and transport potential to move around 34the planet, representing a potential concern at the global level.

35Three complementary information blocks have been analysed for assessing the bioaccumulation and 36biomagnification potential of endosulfan and its degradation products: the screening assessment based on 37physical-chemical properties; the analysis of experimental data, including bioconcentration, bioaccumulation 38and toxicokinetic studies; and the analysis of field collected information.

39The reported log Kow for α- and β-isomers and endosulfan sulfate range between 3 and 4.8. These values 40indicate potential for bioconcentration in aquatic organisms, although are below the screening trigger of the

41Stockholm Convention. Recently, the role of the octanol/air partition coefficient K oa for the screening 42assessment of the biomagnification potential of POPs in terrestrial food chains is receiving significant

43attention. Although there are no specific screening thresholds for the Koa, the authors suggest that organic 44chemicals with a log Kow higher than 2 and a log Koa higher than 6 have an inherent biomagnification potential 45in air-breathing organisms of terrestrial, marine mammalian, and human food chains. Endosulfan clearly falls 46within this category.

2 1The bioconcentration potential of endosulfan in aquatic organisms is confirmed by experimental data. The 2validated BCF values range between 1000 and 3000 for fish,; from 12 to 600 for aquatic invertebrates; and up 3to 3278 in algae. These values, measured in conventional studies, are in line with those expected from the 4Kow, indicating a clear bioconcentration potential but below the screening trigger of 5000. However, due to the 5complex degradation and metabolism pattern of endosulfan, the potential for bioconcentration requires 6additional considerations.

7The data obtained in the estuarine and freshwater microcosm experiments confirms that the assessment of 8parent and metabolite bioconcentration is particularly relevant. In the short-term estuarine experiment, the 9authors suggest BAFs between 375 and 1776 for total (α-, β- and endosulfan sulfate); but BAFs over 5000 10could be obtained for α-endosulfan based on the concentrations measured at the end of the experiment. An 11outdoor aquatic microcosms study estimated bioaccumulation factors of about 1000, based on total 12radioactivity but up to 5000 for endosulfan sulfate. A similar situation is observed in the dietary exposure 13experiments with aquatic organisms. The initial “standard” assessment indicates a low bioaccumulation from 14food in cladocerans exposed to contaminated algae and in fish exposed to contaminated food. However, an in- 15depth analysis of the results in terms of the comparative assessment of the long-term toxicokinetics of 16endosulfan and its degradation products reveal some concerns, for example, the endosulfan concentrations in 17the fish exposed to endosulfan in the diet were low but remained unchanged during the whole depuration 18phase.

19The biomagnification potential of endosulfan has been recently associated with its high K oa, and model 20estimations, based on measured concentration of key elements from remote Arctic food chains, indicates a 21significant biomagnification of endosulfan in terrestrial ecosystems.

22This complex situation has been confirmed by the presence of endosulfan in biota from remote areas. Most 23studies include α- and β-endosulfan, and in some cases, endosulfan sulfate is also measured. Other endosulfan 24metabolites are only rarely quantified. The presence of endosulfan in biota including top predators has been 25confirmed for situations representing medium range transport; potential for long range transport, including 26atmospheric transfer and deposition at high altitude mountain areas; and in remote areas, far away from 27intensive use areas, in particular, the Arctic and the Antarctic.

28Regarding the potential of endosulfan for producing adverse effects, the toxicity and ecotoxicity of this 29pesticide is well documented. Endosulfan is highly toxic for humans and for most animal groups, showing 30both acute and chronic effects at relatively low exposure levels. Acute lethal poisoning in humans and clear 31environmental effects on aquatic and terrestrial communities have been observed under standard use 32conditions when the risk mitigation measures have not been followed. A large number of countries have 33found that endosulfan poses unacceptable risks, or has caused unacceptable harm, to human health and the 34environment, and as a result have banned or severely restricted it.

35Regarding environmental exposure, the potential risk of endosulfan is not limited to zones in the vicinity of 36the areas with extensive use. Concentrations of potential concern have been observed in areas at significant 37distances, due to medium-range atmospheric transport.

38Endosulfan is still in use in several regions and the medium-term transport studies were conducted in world 39regions with a significant use of endosulfan at the time of the study. A particularly relevant trend is observed 40in the North American studies on national parks. Regional differences in use patterns explain the observed 41variable relevance for currently use pesticides in parks located in the temperate region, at medium distances 42from the agricultural areas. However, for those parks located in the northern part of the USA, endosulfan 43acquires a prevalent role, and is the most relevant pesticide in use for the observations in Alaska.

44As expected for a currently used pesticide, the concentrations in remote areas tend to be orders of magnitude 45below those predicted/observed in crop areas. However, the assessment of POP and POP-like chemicals 46requires a very specific evaluation, which strongly differs from that employed in the local risk assessment 47employed by regulatory bodies for supporting the registration of pesticides. Regulatory risk assessments for 48pesticides focus on the health and environmental consequences of local episodic exposures, consider the

2 1expected benefits of the application, and the acceptability criteria differ dramatically from those relevant for 2assessing persistent pollutants. POPs have potential for distributing around the world, reaching remote areas, 3and bioconcentrating along the food-chain resulting in a long term exposure of human and wildlife 4populations. Thus, concentrations assumed to be acceptable at the local level in pesticide regulatory 5programmes should not be considered acceptable in a POP assessment. Such an assessment should be 6conducted based on the scientific evidences on the potential adverse effects to human health or to the 7environment resulted by long-range transport of the chemical.

8The long-term concern for chemicals with POP characteristics is associated with its distribution to remote 9areas, which obviously are expected to lead to low but potentially relevant concentrations, followed by 10biologically dominated concentration processes through specific ecological pathways (biomagnification). 11Although traditionally it has been considered that these processes are dominated by the fugacity potential 12associated with very high lipophilicity and very low aquatic solubility, it is now clear that there are other 13mechanisms and routes which may lead to equivalent health and environmental concerns, as demonstrated for 14other POP candidates such as PFOS or HCH isomers. In the particular case of endosulfan, to the relevant 15although limited bioconcentration potential in water-respiring organisms, two additional concerns should be 16added: first, the potential for biomagnification in food chains constituted by air-breathing organisms; second, 17the concern on the long-term consequences of a number of metabolites which maintain the basic chemical 18structure of endosulfan.

20Finally, the role of endosulfan metabolites other than endosulfan sulfate has received limited attention. 21Endosulfan lactone has the same chronic NOEC value as the parent endosulfan isomers. The lactone is 22produced from the degradation of the carboxylic acid and/or the hydroxyether. If the toxicity of each 23metabolite is integrated into the degradation/metabolism process, the result is a biphasic curve. The initial 24degradation step, to endosulfan sulfate, increases the bioaccumulation potential and maintains or slightly 25reduces the toxicity; the further degradation steps provoke a clear reduction in the toxicity and 26bioaccumulation potential, but then further steps, with the formation of the lactone, increase again the toxicity 27and the bioaccumulation potential.

284. Concluding statement

29Endosulfan has been banned or heavily restricted in a large number of countries but it is still extensively used 30in different regions around the world.

31Endosulfan has been reported as a widely distributed pesticide in the atmosphere of northern Polar Regions. 32Concentrations of endosulfan (isomers unspecified) from Arctic air monitoring stations increased from early 33to mid-1993 and remained at that level through the end of 1997. Unlike most other organochlorine pesticides 34that have decreased over time, average concentrations of endosulfan in the Arctic have not changed 35significantly during the last five years.

36The rapid field dissipation of the endosulfan isomers is related to volatility and is associated with atmospheric 37medium and long-range transport. The persistence and bioaccumulation potential are confirmed through the 38combination of experimental data, models and monitoring results. Endosulfan is highly toxic to humans and 39wildlife. Based on the inherent properties, and given the widespread occurrence in environmental 40compartments and biota in remote areas, which compared to toxicological values suggest a significant risk for 41adverse effects, together with the uncertainty associated to the insufficiently understood role of the 42metabolites which maintain the endosulfan chemical structure, it is concluded that endosulfan is likely, as a 43result of its long-range environmental transport, to lead to significant adverse human health and 44environmental effects, such that global action is warranted.

25 References

3Alonso E, Tapie N, Budzinski H, Leménach K, Peluhet L, Tarazona JV. 2008. A model for estimating the 4potential biomagnification of chemicals in a generic food web: preliminary development. Environ Sci Pollut 5Res Int.;15(1):31-40.

6ATSDR (Agency for Toxic Substances and Disease Register). Toxicological Profile for Endosulfan, 7September 2000. Available at: http://www.atsdr.cdc.gov/toxprofiles/tp41.pdf

8Bajpayee M, Pandey AK, Zaidi S, Musarrat J, Parmar D, Mathur N, Seth PK, Dhawan A. 2006. DNA damage 9and mutagenicity induced by endosulfan and its metabolites. Environ Mol Mutagen. Dec; 47(9):682-92

10Barrie, L.A., D. Gregor, B. Hargrave, R. Lake, D. Muir, R. Shearer, B. Tracey, T. Bidleman. 1992. Arctic 11contaminants: sources, occurrence and pathways. Sci. Tot. Environ. 122, 1-74

12Bengston Nash SM, Poulsen AH, Kawaguchi S, Vetter W, Schlabach M. 2008. Persistent organohalogen 13contaminant burdens in Antarctic krill (Euphausia superba) from the eastern Antarctic sector: A baseline 14study. Sci Total Environ 407(1):304-14.

15Bentzen TW, Muir DCG, Amstrup SC, O’Hara TM. 2008. Organohalogen concentrations in blood and 16adipose tissue of Southern Beaufort Sea polar bears. Sci Total Environ 406:352-67.

17Berntssen MH, Glover CN, Robb DH, Jakobsen JV, Petri D. 2008. Accumulation and elimination kinetics of 18dietary endosulfan in Atlantic salmon (Salmo salar). Aquat Toxicol.; 86(1):104-11.

19Bidleman, D.F., Cotham, W.E., Addison, R.F., Zinck, M.E. 1992. Organic contaminants in the Northwest 20Atlantic atmosphere at Sable Island, Nova Scotia 1988-89. Chemosphere 24, 1389-1412

21Bidleman, T.F., E.J. Christensen, W.N. Billings. 1981. Atmospheric transport of organochlorines in the North 22Atlantic gyre. J. of Marine Research (39), 443-464

23Bidleman, T.F., R.L. Falconer, M.D. Walla. 1995. Toxaphene and other organochlorine compounds in air and 24water at Resolute Bay, N.W.T. Canada..Sci. Tot. Environ. 160/161, 55-63

25Blais, J.M., D.W: Schindler, D.C.G. Muir, L.E. Kimpe, D.B. Donals, B. Rosenberg. 1998 Accumulation of 26Persistent Organochlorine Compounds in mountains of Western Canada. Nature 395: 585-588

27Bostanian Noubar J; Akalach Mohammed 2004. The contact toxicity of indoxacarb and five other insecticides 28to Orius insidiosus (Hemiptera: Anthocoridae) and Aphidius colemani (Hymenoptera: Braconidae), 29beneficials used in the greenhouse industry. Pest management science; 60(12):1231-6

30Brown, Trevor N., and Frank Wania. 2008. Screening Chemicals for the Potential to be Persistent Organic 31Pollutants: A Case Study of Arctic Contaminants. Environ. Sci. Technol., 42 (14), 5202-5209

32Brun G.L. Howell G.D. H. J. O’Neil,. 1991. Spatial and temporal patterns of ogganic contaminants in wet 33precipitation in Atlantic Canada. Environ. Sci. Techno. 27 : 910-914.

2 1Brunelli E, Bernabò I, Berg C, Lundstedt-Enkel K, Bonacci A, Tripepi S. 2009. Environmentally relevant 2concentrations of endosulfan impair development, metamorphosis and behaviour in Bufo bufo tadpoles. 3Aquat Toxicol.; 91(2):135-42

4Buerkle 2003 Endosulfan -Evaluation of estimation of half file in atmosphere MRID 46029902 END. 5CropLife submission.

6California Department of Pesticide Regulation, Environmental Hazard Assessment Program (EHAP), United 7States Geological Survey (USGS), and the Central Valley Regional Water Quality Control Board carried out 8pesticide monitoring studies for surface water (CDPR 2000).

9Carrera G., P., Fernandez, J.O. Grimalt, M. Ventura, L. Camarero,, J. Catalan, U. Nickus, H. Thies, R. 10Psenner. 2002. Atmospheric deposition of organochlorine compounds to remote high mountain lakes of 11Europe. Environ. Sc. Technol. 36: 2581-2588.

12Chan MP, Morisawa S, Nakayama A, Kawamoto Y, Sugimoto M, Yoneda M. 2006. A physiologically based 13pharmacokinetic model for endosulfan in the male Sprague-Dawley rats. Environ Toxicol.; 21(5):464-78.

14Chatterjee S, Kumar V, Majumder CB, Roy P. 2008. Screening of some anti-progestin endocrine disruptors 15using a recombinant yeast based in vitro bioassay. Toxicol In Vitro.;22(3):788-98

16Chernyak S.M., C.P. Rice, L.L. McConnell. 1996. Evidence of currently-used pesticides in air, ice, fog, 17seawater and surface. microlayer in the Bering and Chukchi Seas. Marine Pollution Bulletin 22 (5), 410-419

18Chowdhury AG, Das C, Kole RK, Banerjee H, Bhattacharyya A. 2007. Residual fate and persistence of 19endosulfan (50 WDG) in Bengal gram (Cicer arietinum). Environ Monit Assess.; 132(1-3):467-73.

20Ciglasch H, Busche J, Amelung W, Totrakool S, Kaupenjohann M. 2006. Insecticide dissipation after 21repeated field application to a Northern Thailand Ultisol. J Agric Food Chem.; 54(22):8551-9

22Ciglasch H, Busche J, Amelung W, Totrakool S, Kaupenjohann M. 2008. Field aging of insecticides after 23repeated application to a northern Thailand ultisol. J Agric Food Chem.; 56(20):9555-62.

24De Wit, C.A., A.T. Fisk, K.E. Hobbs, D.C.G. Muir. Levels, trends and effects of Persistent Organic Pollutants 25(POPs) in the Arctic environment. 2nd AMAP International Symposium on Environmental Pollution in the 26Arctic, Rovaniemi 1-3 October 2002

27DeLorenzo ME, Taylor LA, Lund SA, Pennington PL, Strozier ED, Fulton MH. 2002. Toxicity and 28bioconcentration potential of the agricultural pesticide endosulfan in phytoplankton and zooplankton. Arch 29Environ Contam Toxicol.;42(2):173-81.

30Drooge van, B.;L., J.O. Grimalt. 2004. Atmospheric semivolatile organochlorine compounds in European 31High-Mountain areas (Central Pyrenees and High Tatra). Environ. Sci. Technol. 38: 3525-3532

32Durukan P, Ozdemir C, Coskun R, Ikizceli I, Esmaoglu A, Kurtoglu S, Guven M. 2009. Experiences with 33endosulfan mass poisoning in rural areas. Eur J Emerg Med.;16(1):53-6.

34EFSA, 2005. Opinion of the Scientific Panel on Contaminants in the Food Chain on a request from the 35Commission related to endosulfan as undesirable substance in animal feed Question N° EFSA-Q-2003-066 36The EFSA Journal (2005) 234, 1 – 31

2 1Export Import Data Bank. Export: Commodity-wise all countries. Commodity 38081018. Endosulfan 2technical. Government of India, Ministry of Commerce & Industry, Department of Commerce, 3http://commerce.nic.in/eidb/Default.asp.

4FAO/WHO (Food and Agriculture Organization/World Health Organization), 1998. Joint FAO/WHO 5Meeting on Pesticide Residues (JMPR). Endosulfan, part II, toxicology. Available at: 6http://www.inchem.org/documents/jmpr/jmpmono/v098pr08.htm

7Franz, T.P., D.J. Gregor, S.J. Eisenreich. 1997. Snow deposition of atmospheric organic chemicals in: Baker, 8J.E. editor. Atmospheric deposition of contaminants to the Great Lakes and coastal waters. Pensacola, FL: 9Society for Environmental Toxicology and Chemistry 73-107

10Gabrielsen G.W., L.B. Knudsen, M. Schlabach. 2005. Organic Pollutants in Northern Fulmars (Fulmarius 11glacialis) from Bjørnøya SPFO-Report 922/2005

12GFEA (German Federal Environment Agency). 2007. Draft Dossier prepared in support of a proposal of 13endosulfan to be considered as a candidate for inclusion in the UN-ECE LRTAP protocol on persistent 14organic pollutants. German Federal Environment Agency. Umweltbundesamt, Berlin. 15 http://www.unece.org/env/popsxg/docs/2004/Dossier_Endosulfan.2004.pdf

16Girón-Pérez MI, Montes-López M, García-Ramírez LA, Romero-Bañuelos CA, Robledo-Marenco ML. 2008. 17Effect of sub-lethal concentrations of endosulfan on phagocytic and hematological parameters in Nile tilapia 18(Oreochromis niloticus). Bull Environ Contam Toxicol.; 80(3):266-9

19Glover CN, Petri D, Tollefsen KE, Jørum N, Handy RD, Berntssen MH. 2007. Assessing the sensitivity of 20Atlantic salmon (Salmo salar) to dietary endosulfan exposure using tissue biochemistry and histology. Aquat 21Toxicol.; 84(3):346-55

22Gregor, D.J., W. Gummer. 1989. Evidence of atmospheric transport and deposition of organochlorine 23pesticides and PCB in Canadian Arctic snow. Environ. Sci. Technol. 23 (5), 561-565

24Hafner, W. D.and Hites, R. A. 2003. Potential sources of pesticides, PCBs, and PAHs to the atmosphere of 25the Great Lakes. Environ. Sci. Technol. 37: 3764-3773.

26Hageman K.J., Simonich S. L., Campbell D.H. Wilson G.R., and D.H. Landers. 2006. Atmospheric 27deposition of current-use and historic use pesticides in snow at National Parks in the Western United States. 28Environ. Sci. and Tech. 40: 3174-3180.

29Halsall, C.J., R. Bailey, G.A. Stern, L.A. Barrie, P. Fellin, D.CG. Muir, B. Rosenberg, F.Ya. Rovinsky, E.Ya. 30Kononov, B. Pastukhov. 1998. Multi-year observations of organohalogen pesticides in the Arctic atmosphere. 31Environmental Pollution 102, 51-62

32Hansen, D.J., G.M. Cripe. Interlaboratory comparison of the Early Life-Stage Test using sheephead minnows 33(Cyprinodon variegates). In: Aquatic Toxicity and Risk Assessment, edited by M.A.

34Harman-Fetcho, J.A., L.L. McConnell, C.P. Rice, and J.E. Baker. 2000. Wet deposition and air–water gas 35exchange of currently used pesticides to a subestuary of the Chesapeake Bay. Environ. Sci. Technol. 3634:1462–1468.

37Harris, M.L.; Van den Heuvel, M.R.; Rouse, J.; Martin, P.A.; Struger, J.; Bishop, C.A.; Takacs, P. Pesticides 38in Ontario: 2000. A Critical Assessment of Potential Toxicity of Agricultural Products to Wildlife, with 39Consideration for Endocrine Disruption. Volume 1: Endosulfan, EBDC fungicides, Dinitroaniline herbicides,

2 11,3-Dichloropropene, Azinphos-methyl, and pesticide mixtures. Technical Report Series No.340. Canadian 2Wildlife Service, Ontario Region.

3Hobbs, K.E., D.C.G. Muir, E.W. Born, R. Dietz, T. Haug, T. Metcalfe, C. Metcalfe, N. Øien. 2003. Levels 4and patterns of persistent organochlorines in minke whale (Balaenoptera acutorostrata) stocks from the North 5Atlantic and European Arctic Environmental Pollution 121 (2), 239-252

6Hoh, E.; Hites, R. A. 2004. Sources of toxaphene and other organochlorine pesticides in North America as 7determined by air measurements and potential source contribution function analyses. Environ. Sci. Technol. 838: 4187-4194.

9Hung H., Halsall C.J., Blanchard P., Li H., Fellin P., Stern G., Rosenberg B. 2002. Temporal trends of 10organochlorine pesticides in the Canadian Arctic atmosphere. Environ Sci Technol. 36:862-868

11Jantunen L.M.Mannd T.F. Bidleman. 1998. Organochlorine Pesticides and Enantiomers of Chiral Pesticides 12in the Arctic Ocean Water. Arch. Environ. Contam. Toxicol. 35 218-228

13Jergentz S, Mugni H, Bonetto C, Schulz R. 2004. Runoff-related endosulfan contamination and aquatic 14macroinvertebrate response in rural basins near Buenos Aires, Argentina. Arch Environ Contam 15Toxicol.;46(3):345-52.

16Jones W. 2002. Degradation of [14C] Endosulfan in two aerobic water/sediment systems. Reference: 17C022921. EU Additional Information Dossier.

18Jones W. 2003. Degradation of [14C] Endosulfan in two aerobic water/sediment systems (under acid 19conditions). Reference: C031060. EU Additional Information Dossier.

20Jonsson, C.M., M.C.F. Toledo. 1993. Bioaccumulation and elimination of endosulfan in the fish Yellow Tetra 21(Hyphessobrycon bifasciatus). Bull. Environ. Contam. Toxicol. 50(4), 572-577.

22Kang HS, Gye MC, Kim MK. 2008. Effects of endosulfan on survival and development of Bombina orientalis 23(Boulenger) embryos. Bull Environ Contam Toxicol.; 81(3):262-5

24Kelly BC, Gobas FAPC. 2003. An arctic terrestrial food-chain bioaccumulation model for persistent organic 25pollutants. Environ Sci Technol 37(13):2966-74.

26Kelly BC, Ikonomou MG, Blair JD, Morin AE, Gobas FAPC. 2007. Food web-specific biomagnification of 27persistent organic pollutants. Science 317:236-9.

28Kelly, Barry C., and Frank A. P. C. Gobas. 2003. An Arctic Terrestrial Food-Chain Bioaccumulation Model 29for Persistent Organic Pollutants. Environ. Sci. Technol., 37 (13), 2966-2974

30Kelly, Barry C., Michael G. Ikonomou, Joel D. Blair, Anne E. Morin, Frank A. P. C. Gobas. 2007. Food web- 31specific biomagnification of persistent organic pollutants. Science, 317, p. 236

32Konoplev, A., P. Fellin, H. Li, P. Blanchrd, H. Hung, D. Samsonov, G. Stern. 2002. Monitoring of POPs in 33Arctic Ambient Air: Initial results from Anderma (Russia) and Preliminary Assessment. 2nd AMAP 34International Symposium on Environmental Pollution in the Arctic, Rovaniemi 1-3 October 2002

35Kumar S, Habib K, Fatma T. 2008. Endosulfan induced biochemical changes in nitrogen-fixing 36cyanobacteria. Sci Total Environ.; 403(1-3):130-8. Epub 2008 Jun 26

2 1LeNoir JS, McConnell LL, Fellers GM, Cahill TM, Seiber JN. 1999. Summertime transport of current-use 2pesticides from California’s Central Valley to the Sierra Nevada Mountain Range, USA. Environ Toxicol 3Chem 18:2715–2722.

4Li J, Zhu T, Wang F, Qiu XH, Lin WL. 2006. Observation of organochlorine pesticides in the air of the Mt. 5Everest region. Ecotoxicol Environ Saf 63(1):33-41.

6Li, Y. F. and R. MacDonald, 2005, Sources and pathways of selected organochlorine pesticides to the Arctic 7and the effect of pathway divergence on HCH trends in biota: A review, the Science of the Total 8Environment, 342, 87-106

9Mackay N, Arnold D. 2005. Evaluation and Interpretation of Environmental Data on Endosulfan in Arctic 10Regions. Draft Report for Bayer CropScience Report Number CEA.107.

11Mackay N, Arnold D. 2005. Evaluation and Interpretation of Environmental Data on Endosulfan in Arctic 12Regions. Draft Report for Bayer CropScience Report Number CEA.107.

13Majewski M.S. and P.D. Capel . 1995. Pesticides in the atmosphere- distribution, Trends, and Governing 14Factors. Ann Arbor Press, Chelsea, USA.

15McConnell, L.L., J.S. Lenoir, S. Datta, and J.N. Seiber. 1998. Wet deposition of currentuse pesticides in the 16Sierra Nevada mountain range, California. Environ. Toxicol. Chem. 17(10), 1908-1916.

17Meakin, S. What´s New with POPs Research in the Arctic Northern Perspectives 26 (1), 6-7 (2000)

18Menone ML, Pesce SF, Díaz MP, Moreno VJ, Wunderlin DA. 2008. Endosulfan induces oxidative stress and 19changes on detoxication enzymes in the aquatic macrophyte Myriophyllum quitense. Phytochemistry; 2069(5):1150-7

21Morris A.D., D.C.G. Muir, K.R. Solomon, C. Teixeira, J. Epp, A.T. Fisk, R. Letcher, and X. Wang. 2008. 22Current-use pesticide bioaccumulation in Canadian Arctic ringed seal (Phoca hispida)food webs. Presented at 23Arctic Change 2008, Quebec, Canada, December 2008.

24Muehlberger, B., Lemke, G. 2004. Endosulfan and metabolites, partition coefficient 1-octanol/water (HPLC- 25method), endosulfan hydroxy carboxylic acid, sodium salt; endosulfan hydroxy ether; endosulfan lactone; 26endosulfan sulfate; endosulfan ether; beta-endosulfan, alpha-endosulfan. Bayer CropScience, Doc. No. 27C042001.

28Muir, Derek C.G., Camilla Teixeira, and Frank Wania. 2004. Empirical and modelling evidence of regional 29atmospheric transport of current-use pesticides. Environmental Toxicology and Chemistry, Vol. 23, No. 10, 30pp. 2421–2432

31Ngabè, B., T.F. Bidleman. 2001. Endosulfan in the Atmosphere, Review and Evaluation. Report for Center of 32Coastal Environmental Health and Biomolecular Research, National Ocean Service, national Oceanic and 33Atmospheric Administration, Charleston, SC 29412, U.S.A. (2001)

34Palma P, Palma VL, Matos C, Fernandes RM, Bohn A, Soares AM, Barbosa IR. 2009. Effects of atrazine and 35endosulfan sulfate on the ecdysteroid system of Daphnia magna. Chemosphere.; 74(5):676-81.

36Patton G.W., Walla M.D. Bidleman T.F. B.T. Hargrave. 1989. Airbrone organochlorines in the Canadian high 37Arctic. Tellus 41 B: 243-255.

2 1Patton, G.W., D.A. Hinckley, M.D. Walla, T.F. Bidleman. 1989. Airborne organochlorines in the Canadian 2High Arctic. Tellus, 41B, 243-255

3Pennington, P.L., DeLorenzo, M.E., Lawton, J.C., Strozier, E.D., Fulton, M.H., and G.I. Scott. 2004. Modular 4Estuarine Mesocosm Validation: Ecotoxicological Assessment of direct effects with a model compound 5endosulfan. J. Exp. Mar. Biol. Ecol. 298: 369-387

6Performance of Chemical & Petrochemical Industry at a Glance (2001-2007). Monitoring and Evaluation 7Division, Department of Chemicals & Petrochemicals, Ministry of Chemicals & Fertilizers, Government of 8India, New Delhi. http://www.chemicals.nic.in/stat0107.pdf.

9Petri D, Glover CN, Ylving S, Kolås K, Fremmersvik G, Waagbø R, Berntssen MH. 2006. Sensitivity of 10Atlantic salmon (Salmo salar) to dietary endosulfan as assessed by haematology, blood biochemistry, and 11growth parameters. Aquat Toxicol.; 80(3):207-16

12Roseneau DG, Becker PR, Vander Pol SS, Day RD, Point D, Simac KS, Moors AJ, Ellisor MB, Pugh RS, 13York GS. 2008. Expanding the Seabird Tissue Archival and Monitoring Project (STAMP) in the North 14Pacific: Geographic Patterns in Contaminant Residues in Seabird Eggs Used in Rural Subsistence Diets. 15North Pacific Research Board Project Final Report (NPRB Project 0534). 16http://doc.nprb.org/web/05_prjs/534_Final%20Report%20(Mar%202008)%20(2).pdf

17Schanne, 2002. [14C]-Endosulfan formulated as emulsifiable concentrate (352g/l endosulfan): outdoor 18aquatic microcosm study of the environmental fate and ecological effects. Springborn Laboratories. 500pp.

19Schimmel, S.C. et al.1977. Acute toxicity to and bioconcentration of endosulfan in estuarine animals. In: 20Aquatic Toxicology and Hazard Evaluation, edited by F.L. Mayer, J.L. Hamelink, 1stSymp. ASTM STP 634, 21Philadelphia (PA), 241-252

22Shen L., F. Wania, F. Lei, Y.D., D.C.G Muir and T. Bidleman. 2005. Atmospheric distribution and long lange 23transport behavior of organochlorine pesticides in north America. Environ. Sci. and Technol. 39: 409-420

24Stoker C, Beldoménico PM, Bosquiazzo VL, Zayas MA, Rey F, Rodríguez H, Muñoz-de-Toro M, Luque EH. 252008. Developmental exposure to endocrine disruptor chemicals alters follicular dynamics and steroid levels 26in Caiman latirostris. Gen Comp Endocrinol.; 156(3):603-12

27Sun P., Basu I., Blanchard P., Backus S.M., Brice K. L., Hulting M.L., Hites R.A. 2003. temporal and spatial 28trends of atmospheric toxic substances near the great lakes: IADN results through 2003. Environment Canada 29and the United States Environmental Protection Agency, Chicago IL

30Sun P., P. Blancharrd, K, B. Kenneth, and R.A. Hites. 2006. Atmospheric organochlorine pesticide 31concentrations near the Great Lakes: temporal and spatial trends. Environ. Sci. and Tech. 40: 6587-6593

32Survey of Chemical Contaminants in Fish, Invertebrates, and Plants Collected in the Vicinity of Tyonek, 33Seldovia, Port Graham, and Nanwalek in Cook Inlet, Alaska. December 2003. Prepared by the U.S. 34Environmental Protection Agency Region 10 Office of Environmental Assessment (EPA 910-R-01-003).

35 http://yosemite.epa.gov/r10/oea.nsf/Risk+Assessment/Cook+Inlet+Seafood+Study

36Tellez-Bañuelos MC, Santerre A, Casas-Solis J, Bravo-Cuellar A, Zaitseva G. 2008. Oxidative stress in 37macrophages from spleen of Nile tilapia (Oreochromis niloticus) exposed to sublethal concentration of 38endosulfan. Fish Shellfish Immunol.

2 1UNEP-POPS-POPRC-END-08-EU-A6.English

2USEPA, 2007. Appendix 1 to 2007 Addendum: Environmental Fate and Ecological Risk Assessment of 3Endosulfan. USEPA, 101pp.

4Varayoud J, Monje L, Bernhardt T, Muñoz-de-Toro M, Luque EH, Ramos JG. 2008. Endosulfan modulates 5estrogen-dependent genes like a non-uterotrophic dose of 17beta-estradiol. Reprod Toxicol.; 26(2):138-45.

6Vig K, Singh DK, Sharma PK. 2006. Endosulfan and quinalphos residues and toxicity to soil microarthropods 7after repeated applications in a field investigation. J Environ Sci Health B.; 41(5):681-92

8Vorkamp K, Riget F, Glasius M, Pecseli M, Lebeuf M, Muir D. 2004. Chlorobenzenes, chlorinated pesticides, 9coplanar chlorobiphenyls and other organochlorine compounds in Greenland biota. Sci Total Environ 331(1- 103):157-75.

11Wania F, Mackay D, 1993a. Global fractionation and cold condensation of low volatile organochlorine 12compounds in polar regions. Ambio 22:10-18

13Watson CS, Bulayeva NN, Wozniak AL, Alyea RA. 2007. Xenoestrogens are potent activators of 14nongenomic estrogenic responses. Steroids.; 72(2): 124–134

15Weber J, Halsall CJ, Muir DC, Teixeira C, Burniston DA, Strachan WM, Hung H, Mackay N, Arnold D, 16Kylin H. 2006. Endosulfan and gamma-HCH in the arctic: an assessment of surface seawater concentrations 17and air-sea exchange. Environ Sci Technol.; 40(24):7570-6

18Wessel N, Rousseau S, Caisey X, Quiniou F, Akcha F. 2007. Investigating the relationship between 19embryotoxic and genotoxic effects of benzo[a]pyrene, 17alpha-ethinylestradiol and endosulfan on Crassostrea 20gigas embryos. Aquat Toxicol.; 85(2):133-42